Author: Dieter Buerssner
Date: 14:11:26 09/20/03
Go up one level in this thread
On September 20, 2003 at 13:37:34, Drexel,Michael wrote: >Yace Paderborn was not able to find the win at depth 25 here. If at all, it would probably need a really a long time (to see a good score). This does not automatically mean (but it could - for example if you tell me, that Bd6 already throws away the win), that it could not win the game. >When I came back it searched only ~130 kN/s. This is a misfeature of Yace (you may call it a bug, and I won't be offended). The nodes counter had an overflow already. The displayed kN/s are wrong, It is no real problem. Yace will use the nodes counter (and similar counters, that can overflow during longer analysis) almost only for displaying stuff. The only exception is, that the estimated rate for polling the input is also dependent on it. But this just means, that it will poll the input more often than intended (and the time loss by this is neglible). >89.Rh7 Kg8 90.Ra7 Bg4 91.Kd6 Kf8 92.Ra8+ Kf7 93.Ra4 > +- (2.63) Depth: 24/50 01:24:56 3834350kN Here knodes/s is 752. >89.Rh7 Kg8 90.Ra7 Bg4 91.Kd6 Kf8 92.Ra8+ Kf7 93.Ra4 Bf5 94.Kc7 > +- (2.63) Depth: 24/50 01:51:00 505476kN Here was the first overflow. Actually there are 2**32/1000 kN searched more here. And so on ... Regards, Dieter
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.